Adoptive transfer of genetically modified T cells shows some advantages over the mobilization of the endogenous T cell repertoire in cancer immunotherapy. The theoretical advantages and technical feasibility of CAR facilitate the development of cancer immunotherapy. The notions that CAR endows T cells antigen specific recognition, activation and proliferation in an MHC independent manner have been consolidated by some pre-clinical studies showing that retargeted T cells can recognize and kill cancer cells expressing tumor associated antigen or specific antigen in vitro and in vivo. Cellular immunotherapy adopting such retargeted T cells has shown significant potential in the treatment of malignant diseases. The mounting data have provided solid support for future clinical application of such therapy in cancers such as leukemia, colorectal colon cancer, and prostate cancer [17–19].
Recent efforts to improve the antitumor efficacy of CAR-based therapy are mainly based on the theory of the two-step T cell activation. Major progress has been made since the introduction of the costimulatory signaling into architecture of CAR. With the in-depth understanding of costimulatory receptors in T cell immune response, several costimulatory molecules were embedded in the CAR and their roles in coordinating antitumor immunity were explored . The observations from other groups and our own have thus far established that the inclusion of costimulatory molecule from B7 receptor family (CD28 or ICOS) results in an increased production of of IFN-γ, TNF-γ, and GM-CSF compared with the CAR with the inclusion of either CD134 or CD137 of TNFR family. CD28 is more potent than other costimulatory molecules with respects to enhanced IL-2 production, improved clonal expansion and persistence of CAR T cells. Finney and colleagues have demonstrated that ICOS in the CAR induces the maximal effect on cell lysis . Thus, we hypothesized that the incorporation of ICOS into CAR favors the antitumor properties of CAR-armed T cells. Our results support our hypotheses: CAR-armed T cells demonstrate efficient killing of tumor cells and abundant Th1 cytokine IFN-γ is released in an EGFRvIII-specific manner. Therefore, our results are consistent with the previous findings and consolidate the notion that the presence of ICOS as an intracellular costimulatory signaling is crucial for enhanced T cell response to tumor cells.
The efficacy of CAR can be affected by many factors including the affinity of the selected scFv, the size of hinge region, the combination of signaling domain(s), the type of modified T cell subsets, etc. The major concern of CAR T cell transfer is the possible recognition of the antigen expressed on normal cells by CAR T cells. Such off-site on-target immune injury causes adverse effects, some of which may be fatal. Therefore, it is important to carefully select the target antigens that are specifically expressed in cancer cells, but not in normal cells. EGFRvIII is a commonly found mutant of EGFR and exclusively expressed in a wide range of cancers. In addition to its tumor specific expression, EGFRvIII is also involved in oncogenic phenotypes and changes the properties of tumorigenicity. Since its discovery, EGFRvIII has become an increasingly attractive molecule for cancer therapy. The EGFRvIII scFv from antibody 3C10 and MR1 was used for CAR construction and the CAR modified T cells demonstrated EGFRvIII-specific tumor cell lysis in vitro and in vivo [21, 22]. Recently, Rosenberg’s group analyzed scFv sequences of seven EGFRvIII specific mAbs and assembled the third generation of chimeric antigen receptor (139-28BBZ CAR). The T cells transduced with retroviral EGFRvIII/CAR have been shown an EGFRvIII-specific cell lytic activity . In this study, lentivirus-mediated transduction enriched the EGFRvIII/CAR+ T cells to about 70%. Functionality assay demonstrated that these redirected T cells exert efficient cytotoxic T cell response in an EGFRvIII specific manner and secret cytokine IFN-γ in an antigen dependent way. The EGFRvIII/CAR engrafted T cells pave a way for antitumor in animal model as well as in clinical settings.
In our study, we used CD3+ T cells, instead of purified CD8+ CTLs only, to investigate the performance of CAR because CD4+ T cells have been shown to augment the function of CD8+ T cells. The results of our EGFRvIII-bearing mouse model demonstrates that CD3+ T cells transduced with EGFRvIII/CAR have significantly higher antitumor activity than the T cells in control groups, which supports the theory that adoptive transfer of mixed populations of antigen-specific CD8+ and CD4+ T cells promotes overall antitumor immunity. With regards to the administration route of T cells, both intratumor injection and venous injection show similar efficacy. For clinical purpose, engraftment of adoptively transferred T cells in host is a major challenge for achieving therapeutic benefit. Ongoing studies are exploring optimal combinations of costimulatory molecules and T cell subsets with long-term cytotoxicity . Also, most of mice in groups of the EGFRvIII/CAR+ T cells and GFP+ T cells suffered from graft versus host diseases (GVHD) to some extent. The symptoms of GVHD usually started 5 weeks after T cells infusion, including less activity, ruffled fur, skin rash, hunched back and weight loss. The lethality due to the transfusion-associated GVHD was rare and there was no difference between mice treated with EGFRvIII/CAR+ T cells and GFP+ T cells with regard to the severity of GVHD.